Distances people Abstract walk for transport South East ...

Distances People Walk for Transport

Author

Burke, Matthew, Brown, AL

Published

2007

Journal Title

Road & Transport Research

Copyright Statement

© 2007 ARRB Ltd. The attached file is reproduced here in accordance with the copyright policyof the publisher. Please refer to the journal's website for access to the definitive, publishedversion.

Downloaded from

http://hdl.handle.net/10072/17867

Link to published version

http://search.informit.com.au/documentSummary;dn=880016006301032;res=IELENG

Griffith Research Online

https://research-repository.griffith.edu.au

16

Distances people walk for transport

Vol 16 No 3 September 2007 Road & Transport Research

Distances peoplewalk for transport

Matthew Burke and A.L. Brown

Refereed PaperThis paper has been critically reviewed by at least tworecognised experts in the field.

Originally submitted: July 2007

AbstractWe present detailed information on the distancespeople walk for transport purposes in Brisbane,Australia. The data is derived from the South EastQueensland Travel Survey 2003-04 – Brisbane StatisticalDivision, a household travel survey providinginformation on the weekday travel of 10 931respondents, weighted and expanded for a citypopulation of 1 615 579 persons. The vast majority ofnon-motorised travel recorded was walking fortransport, whether this comprised trips using thewalking mode only, or whether it was walking madeto and from public transport. We report the fulldistributions of the distances walked for transportfrom homes to other places, as well as the walk travelmade between places other than homes. Single-modewalk trips are longer than the walk trip stages madeto and from public transport, both in terms of distanceand time. However, there are more than twice as manywalk trip stages made to and from public transport.The median and 85th percentile distances peoplewalk from home to all other places using the walkmode only are 780 m and 1.45 km respectively; fromhome to all public transport stops, 600 m and 1.30 km;and from public transport stops to end destinations,470 m and 1.09 km.

17



INTRODUCTIONHow far do people walk for transport purposes incities? We present, in this paper, detailed informationon walking distances in Brisbane derived fromhousehold travel survey data. Given the general lackof attention to this form of urban travel, our interest isin providing quantitative information potentially ofuse for policy and planning in both transport andhealth fields.

Walking comprises the bulk of non-motorised travelin many cities (Pucher and Dijkstra 2003; Vivier 2001).Walking can be for transport purposes (so called activetransport) made to access destinations or to accesspublic transport en route to destinations. It is thedistances travelled when walking for transport that isthe focus of this paper. Walking may also be forexercise, sport or recreation – when accessing adestination is not the primary purpose of a trip (Litman2003; Tudor-Locke et al. 2005). Walking is of increasinginterest given the decline of physical activity observedin Western populations (Berrigan et al. 2006; Frank2000).

There are considerable issues in measuring walktravel and in comparing it across cities. Householdtravel surveys (HTSs), the main method of datacollection on travel behaviour, are primarily directedat modes other than walking, and methods used inthese surveys differ across jurisdictions. As a result,there are a number of conceptual and definitionalproblems inherent in any analysis of HTS walkingdata, and comparisons across survey datasets aredifficult. For example, trip distances may be estimatedusing respondent recall alone or may be derived usingthe city’s street network recorded in geographicalinformation systems (GIS). Varying definitions orsurvey methodologies used in different surveys canalso lead to some forms of walking being excluded orunder-reported – recreational walking or walking toand from motor vehicles, for example. The way multi-stage trips are handled also differs. In addition, thereis no consistent approach to reporting the amount andcharacteristics of walking travel (other than perhapsa walking mode share). There is not alwaysrecognition that observed walk trip distances relatenot just to traveller’s preferences, but also to thedistribution of available destinations within the cityand the density of catchments. Finally, much of theavailable literature on this topic is from the US which,given the low mode shares for walking in US cities incomparison to European, Asian and Australasian

cities (see Vivier 2001), may have limited relevance tocities elsewhere.

There has been some attention in the literature to theinfluence of distance (or proximity) on trip generationrates and mode choices for walking (i.e. Krizek andJohnson 2006; Polzin and Maggio 2007) as well as towalk trip distances travelled to particulardestinations. For instance, Hsiao et al. (1997:51) used1990 Orange County Transportation Authority datato show that persons walked on average 0.29 miles(0.47 km) per trip to or from public transport. Plaut(2005:351) used data from the 2001 American HousingSurvey to determine how far people walked fromhomes to workplaces as part of single-mode trips.These US travellers walked on average only 0.20 miles(0.32 km) with a standard deviation of 0.53 miles(0.85 km). For all trip purposes, Land Transport NZ(2004) found more than half of all walking trip stagesmade in New Zealand were less than 500 m.

In Australia attention has primarily been given towalking distances to and from public transport, inpart to assist with public transport stop spacing andland use design. Ker and Ginn (2003:75-76,79) havereported on intercept surveys of commuters at fivemostly outer-suburban rail stations in Perth to showthat walking distances to rail services were muchgreater than the 400 m to 800 m ‘rules of thumb’ usedby transport and land use planners. Wallace (2006)found similar results for walking distances to threebusway stations in Brisbane. Using HTS data from theperiod 1997-2005, the NSW Transport and PopulationData Centre (2006:3-4) found persons walked, onaverage, 700 m from homes to train services in Sydney,and those walking from trains to their end destinationswalked 600 m.

While aware of the limitations of household travelsurvey (HTS) data analysis, this paper examines indetail the data on walking in Brisbane. We focussolely on walking and report the full distributions ofwalking trip distances to the most commondestinations. Throughout the text we also refer to themedian and 85th percentile trip distances. In thispaper there is no disaggregation by demographiccharacteristics or by sub-regional variation ofpopulations or available destinations.

18



METHODThe most recent HTS data for Brisbane was used forthis work – the South East Queensland Travel Survey2003-04 – Brisbane Statistical Division (SEQTS). Thesurvey used a multi-stage, variable-proportion,clustered sampling of households within CensusCollection Districts (CCDs) in 11 sub-regions. TheSEQTS achieved a response rate of 60% and obtainedinformation on the travel behaviour of 10 931respondents living in 4057 households.

Diaries were completed by respondents aged 5 andover, with diaries reconstructed from other householddiaries for children aged 0-4. The respondentscompleted single weekday travel diaries during theperiods October-December 2003 and February-March2004. All trips made by respondents were recorded,with each trip divided into trip stages (for example, apublic transport trip from home to school may involvethree stages: walk stage to the public transport stop,the public transport stage, and the final walk stagefrom the public transport to the school).

In total 41 110 trip stages were recorded for the surveysample in the 35 960 trips that they reported. Therelatively low average of 1.14 trip stages per trip for thecity relates to both the low mode share for publictransport in the city, compared to European cities, andthe definitions and methods used in the SEQTS. Inparticular, walking trip stages made to a motor vehiclewere generally ignored, whilst those made from amotor vehicle were recorded separately and only interms of their travel time, as is discussed furtherbelow. The SEQTS did not use trip-stage enrichmentprocesses such as those developed by Chalasani (2005)that attempt to more fully impute missing walk tripstages.

The exact route travelled by respondents was notcaptured and trip distances were calculated usinggeographic information systems that determined theshortest path possible on the available street and pathnetwork. Motor vehicles were defined in the survey aseither a car, 4WD, van, or truck (Queensland Transportet al. 2005:87-89,133-134).

To account for non-reporting, weightings for bothnon-response and selection bias (derived from

household characteristics and Australian Bureau ofStatistics 2001 census data for the areas surveyed)were included within the SEQTS data set. It was foundthat people in rented, multi-unit dwellings were under-represented in the sample, as were males and peopleaged 15-29 or over 75. The weighting processesgenerally assumed that, for each demographic group,non-respondents to the survey travel in similar waysto respondents. The weightings were applied to thesample results to estimate the active travel parametersfor the city population of 1 615 579 persons(Queensland Transport et al. 2005:73-82). It ispopulation data that is reported in this paper.

RESULTS

Mode ShareMode share is most often reported from HTS databased on the mode of the trip – with multi-stage tripsascribed a mode according to the following hierarchy:public transport modes first, the motor vehicle next,then the bicycle, with walking the lowest in thehierarchy (Queensland Transport et al. 2005:85-86).In such reporting, trips that are undertaken solely bywalking are the only walking trips identified in themode share results. Mode share for Brisbane from theSEQTS, estimated in this manner, is shown in Figure1a – walking represents just over 10% of all trips. Analternative way of looking at mode share is theproportion of all trip stages undertaken using eachmode, and Brisbane results for trip stages are shownin Figure 1b. Over 20% of all trip stages are walked.This different way of looking at the same travel datahighlights the importance of walking within the city,whether as walking only trips or walking as part oftrips that include other modes.

Further, the walking reported in the SEQTS is largelywalking for transport (travel to destinations) ratherthan walking primarily for exercise, sport andrecreation purposes (the latter walking is coded in theSEQTS as being for walking the dog, or for exercise,sport and recreation and not having a destinationwhere such activities occur). In the SEQTS data set,97% of the walking trip stages were walking fortransport. In terms of distance, this represented 96%of the total kilometres walked in Brisbane, with only4% for exercise, sport and recreation purposes.1

1 Comparison with data obtained on walking for transport vs. walking for exercise, sport and recreation in NSW by Cole et al.(2006) suggests that the walking for exercise, sport and recreation is significantly under-reported in the SEQTS (Burke andBrown 2007) – and this is an issue likely to be found in all HTS data sets.

19



Walking for transportWalking for transport was disaggregated accordingto the origins and destinations of each trip, not tripstage, ensuring it is categorised by the place the travellereventually reached at the end of a trip. This travel wascategorised as being from a person’s home to otherplaces, from other places to the person’s home, ortravel between other places. Of the total kilometreswalked for transport purposes in the city, around 36%was made from respondent’s homes to other places,25% was made from other places to homes, and 38%was travel between other places. The differencebetween the travel from homes and the travel to homeswas partly due to differences in the rates at whichpeople walk to public transport. A higher proportionof persons walked from public transport to their homethan walked to public transport from their home, withpersons often riding as passengers in motorcars toaccess public transport. However, on the basis thattravel from homes by walking, and travel to homes bywalking is largely similar but in the reverse direction,we report home-based data only in terms of the travelfrom homes to other places. Walking trips notassociated with the home, solely between other places,are considered separately, below.

For travel from homes to other places, walking isfound in one of three types of trip:

Single-stage, single-mode:

• Walk trips made directly to destinations (single-mode walk trips).

Multi-stage, multi-mode:

• Walk trip stages made to and from public transportstop as part of public transport trips.

• Walk trip stages made to or from a motor vehicleas part of vehicular trips.

The last of these, to and from a motor vehicle, isconsidered separately below.

Table 1 shows a summary of the trip stages made forpersons walking from homes to other places. Single-stage walk trips are longer than the walk trip stagesmade to and from public transport, both in distance(mean = 940 m vs. 640 m) and time (mean = 12.9 minsvs. 7.9 mins). However, there are more than twice asmany walk trip stages made to and from publictransport, such that around 60% of both the totaldistance and time walked is associated with makinga public transport trip, with only some 40% as single-stage, single-mode, walk trips.

Burke and Brown (2007) have analysed the destinationof trips made from home to other places in the SEQTS,

Figure 1Mode shares for trips vs. mode shares fortrip stages

20



Table 2Destinations of trips made from home that include walking for some or all of the trip

Destinations No. of single-stage No. of multi-stagetrips to this destination trips to this destination

(per 1000 persons) (per 1000 persons)by walking that include walking

Primary school 26.0 4.8Shop 22.7 11.8Other1 15.7 11.7Workplaces 9.2 51.6Someone else’s home 8.2 2.2Secondary school 6.3 17.9Pre-schools & childcare 3.8 0.1Restaurants & cafes 2.1 2.2Universities & TAFEs 1.8 13.6

Total 95.8 116.1

1 Includes all destinations not otherwise referred to in the table (e.g. health services, other personal services, places for exercise,sport or recreation, places for entertainment, libraries, airports and petrol stations)

and a summary of these destinations is shown in Table2. Primary schools and shops are the predominantdestinations for walk-only trips and, to a lesser extent,workplaces and someone else’s home. For multi-stagemulti-modal trips that include walking, thepredominant destinations are workplaces, secondaryschools, universities and TAFEs2, and shops.

When considering all travel (travel to/from homesand travel made solely between other places) publictransport-related walking represents 51% of the total

kilometres walked, whereas single-mode walk tripsrepresent 49%.

Distances walked for single-stage walktripsThe distances travelled for all single-stage walkingtrips made from home to other places are shown inFigure 2a. The most common destinations for thesetrips were primary schools (27% of the kilometreswalked to all destinations in single-stage trips), shops

Table 1Mean travel times and distances for walking trip stages made from home to other places.

No. of % of trip % of % of Mean Std. Mean Std.trip stages kilometres minutes distance deviation time per deviation

stages walked walked per trip trip stage stage (km) (min)

Single stage 154 809 31% 38% 41% 0.94 0.79 12.9 9.48walk trips

Walk trip stages 349 982 69% 62% 59% 0.67 0.58 8.23 6.71to/from publictransport

Total 504 791 100% 100% 100% 0.75 0.66 9.66 7.96

2 Technical and Further Education colleges (TAFEs) are higher education providers in Australia.

21



Figure 2aDistances walked for single-stage trips from home to all otherplaces

(21%) and workplaces (11%). No other destinationsrepresented more than 10% of the total distance peoplewalked for these types of trips. The distributions of thedistances walked to these destinations are shown inFigure 2b.

The median distance walked from home to all otherplaces was 780 m (85th percentile = 1.45 km). Distanceswalked to shops (median = 680 m; 85th percentile=1.24 km) and primary schools (median = 790 m; 85th

Figure 2bDistances walked for single-stage trips from home to shops, primary schools and workplaces

percentile =1.34 km) are less than to workplaces(median = 1.04 km; 85th percentile = 1.85 km).

Distances walked at either end of thepublic transport tripThe most common destinations for public transporttrips made from home were to workplaces (representing46% of the kilometres walked for these trips to alldestinations), secondary schools (16%), universitiesand TAFEs (12%) and shops (11%). No otherdestinations represented more than 10% of thekilometres walked for these trips.

Distances walked from homes to any public transportnode (bus stops, train stations and ferry terminals) forall public transport trips to all destinations, are shownin Figure 3a. This is disaggregated by public transportstop type in Figure 3b. The median distance walkedfrom home to all public transport stops is 600 m (85th

percentile = 1.30 km), though persons travel muchless distance to bus stops (median = 440 m; 85th

percentile = 1.07 km) than they do to train stations(median = 890 m; 85th percentile = 1.57 km) or ferryterminals (median = 890m; 85th percentile = 1.54km).The difference in the median walking distances to busstops and train stations, more than 400 m, isstatistically significant (Mann WhitneyU = 969 086 782; p < 0.001). There is also significantdifference in the variances of the travel distancedistributions for bus stops and train stations (F = 1530;p < 0.001). A greater proportion of travellers walkvery short distances (<500 m) to bus stops than to

22



train stations— an outcome of the greater number andavailability of bus stops compared to train stationswithin Brisbane.

Distances walked at the end of these trips, from publictransport nodes to all destinations, are shown inFigure 4a, and disaggregated by public transport stoptype in Figure 4b, and for particular destination typesin Figure 4c. The median distance travelled frompublic transport to end destinations was slightly less

(470 m) than the median distance travelled from homesto public transport (600 m), a difference that isstatistically significant (Mann Whitney U =12 209 166 116; p < 0.001). Persons walking from busstops to end destinations (median = 330 m; 85th

percentile = 850 m) travelled lesser distances thanthose walking from train stations (median = 620 m;85th percentile = 1.32 km), which is also statisticallysignificant (Mann Whitney U = 3 431 321 892;p < 0.001), and again there is a difference in thevariances of the walking distances for the two modes(F = 3996; p < 0.001). Around 40% of persons walkingfrom bus stops to end destinations travel less than400 m, compared to only 15% walking less than 400 mfrom train stations.

Persons also walk, on average, further from publictransport stops to secondary schools (median = 570 m;85th percentile = 1.38 km), universities and TAFEs(median = 590 m; 85th percentile = 1.53 km) andworkplaces (median = 490 m; 85th percentile = 990 m)than they do to shops (median = 310 m; 85th percentile= 760 m) – see Figure 4c. This in part relates to thegreater use of buses than trains for shopping trips.

Distances walked between destinationsother than homesTravel made by persons solely between other places(not to or from their homes) comprised around 25% ofthe total walking kilometres for transport purposes.Some 72% of the kilometres walked between otherplaces was undertaken by single-stage walking trips,and only 28% was made by multi-stage trips involving

Figure 3aDistances walked from home to all public transportnodes as part of trips to all other places

Figure 3bDistances walked from home to bus stops, ferry terminals and train stations as part of trips to other places

23



Figure 4b Distances walked from bus stops, ferry terminals and train stations to end destinations as part of trips from home to other places

public transport. Closer investigation revealed thatmuch of the public transport related travel was part oflonger trip chains involving travel to or from homes.

Much of the single-stage walking travel was madebetween specific origins and destinations, namely,between workplaces and shops (12.3% of thekilometres walked for all origin and destination pairs),from one shop to another shop (12.1%) and between

workplaces and restaurants or cafes (11.9%). No otherorigin and destination pairs were responsible formore than 10% of the kilometres walked for this travel.Distances walked for all single-stage walk trips madebetween other places are shown in Figure 5a, anddisaggregated for the most important origin anddestination pairs in Figure 5b.

The median distance walked for single-stage walkingtrips between other places was 430 m (85th percentile= 1.09 km). On average, people walked slightly furtherfrom shop to shop (median = 410 m; 85th percentile =1.18 km) than they did when travelling betweenworkplaces and shops (median = 390 m; 85th percentile= 910 m) and between workplaces and restaurants orcafes (median = 330 m; 85th percentile = 870 m). Notethat the distance walked between shops is influencedby the means by which this travel was recorded in theSEQTS. The travel diaries used emphasise travelbetween buildings, not between individual shops–such that travel within shopping malls from shop toshop is generally not reported.

Distances walked from motor vehiclesto destinationsThe walk to access motor vehicles is trivial. Informationon the walk from where motor vehicles were parked todestinations was collected in the SEQTS, from driversand passengers, but only in terms of estimated minutesof walking from where the motor vehicle was parkedto the end destination. This data suffers from the‘lumpiness’ of all such time estimates (whereby

Figure 4aDistances walked from public transport nodes to enddestinations as part of trips from home to other places

24



Figure 5aDistances walked for single-stage trips made betweendestinations other than homes

Figure 4cDistanceswalked frompublictransportnodes toshops,secondaryschools,universities/TAFEs andworkplaces, aspart of tripsfrom home toother places

respondents tend to indicate they walked either oneminute, two minutes or in multiples of five minutes).While recognising such weaknesses, these times havebeen converted to approximate distance travelled,using a walking speed of 1.26 m/s (75 m/min)calculated as the average speed of all other walkingfor transport made within the SEQTS. The distributionof these walking distances is shown in Figure 6.

Persons walking at least one minute travelled a mediandistance of 75 metres (1 min) from their vehicles todestinations (mean = 150 m or 2 min; 85th percentile= 225 m or 3 min). The number of walk trip stagesmade on exiting a motor vehicle is large – not surprisinggiven motor vehicles account for 79% of all trips, and70% of all trip stages, made in Brisbane, but these walktrip stages are very much shorter than those madeeither to or from public transport, or as part of single-mode walk trips.

25



FUNCTIONS FITTED TO THEDISTRIBUTIONSThe walking distance distributions were fitted withvarious functions using EasyFit software, which testsgoodness-of-fit of a number of distribution functions.It was found that these distributions were consistentlybest fitted by the gamma distribution. Taylor andYoung (1988:73-74) note that the gamma distributionis useful in describing transport data which cannot bereadily characterised by either a negative exponential

distribution or a gaussian distribution. The fitting ofgamma functions to the walking-distancedistributions is not intended to suggest additionalinsight into walking behaviour is provided by thegamma functions – the purpose has to been toefficiently mathematically describe the distributions.The gamma function features two parameters – a scaleparameter (±) and a shape parameter (²). Methods tocalculate parameter values and a full explanation ofthe gamma distribution are found in Johnson et al.(1994: 337-414).

Probability density functions for the gammadistributions that approximate the measured traveldistributions were calculated. Figure 7a shows thegamma function fitted to the walking distancedistribution previously identified for all single-stagewalking trips made from home to other places (Figure2a). Figure 7b shows the same for walking trips madefrom train stations to all other destinations (Figure 4b).

Anderson-Darling (A-D) tests examine how well theobserved walking distance data are estimated bygamma functions. The A-D test is a robust non-parametric test based on the maximum distancebetween the observed and hypothesised curves and isa modification of the Kolmogorov-Smirnov (K-S) test.The A-D test gives more weight to the tails of thedistributions than the K-S test and may be consideredas conservative in situations where the functionparameters are estimated from the data (see Stephens(1974) or NIST and SEMATECH (2004)). Table 3 showsthe parameters for the different gamma functions thathave been fitted to each of the measured walkingdistance distributions reported in the paper, and the

Figure 5bDistances walked for single-stage trips made between shops, between workplaces and shops, and between workplaces andrestaurants/cafes

Figure 6Distances walked when exiting a motor vehicle and walkingto an end destination, for trips made from home to otherplaces

26



Figure 7aGamma function fitted to the distribution of walking tripdistances for single-stage trips from home to all other places

Figure 7bGamma function fitted to the distribution of walking tripdistances from train stations to end destinations as part oftrips from home to other places

A-D test statistic for whether that measured sample ofwalking distances could have come from a populationof distances with that gamma distribution. Despitethe conservative nature of the test, the gammafunctions fit the distributions well in most cases,though trip types with smaller numbers of observationswere less likely to have walking distance distributionsthat closely follow the gamma function.

DISCUSSION AND CONCLUSIONSWalking for transport is shown to be an importantcomponent of travel in Brisbane. As one in five tripstages in Brisbane involves walking, then moreanalysis of walking, more understanding of its

parameters, and more planning and design toencourage it, is warranted.

For trips using the walk mode only, people are willingto walk sizeable distances. The median and 85th

percentile distances people walk from home to allother places using the walk mode only are 780 m and1.45k m respectively. However, distances for walktrip stages made to and from public transport are alsosubstantial. Median and 85th percentile walkdistances from home to all public transport stops are600 m and 1.30 km; and from public transport stops toend destinations, 470 m and 1.09 km. There are alsomore than twice as many walk trip stages made to orfrom public transport as there are single-mode walktrips. The result is that, in total, persons who are usingpublic transport in Brisbane each weekday (206 000persons, 12.8% of the population) are, on average,walking more than 2.3 km and over 28 minutes to andfrom public transport. While it could be argued thatsuch walking may not be the same as deliberateexercise, any more than any other sort of ‘incidental’walking is, it could be noted in passing that, for thesetravellers, public transport-related walking almostmeets the Australian daily minimum physical activityrecommendation of 30 minutes (Egger et al. 1999).

Though comparisons across datasets are problematicthe results are in line with Ker and Ginn’s (2003) andWallace’s (2006) findings regarding walking tripdistances to public transport. Walking trip distancesare much further than the 400 m to 800 m rule-of-thumb proximities (‘¼ to ½ mile’) for walkingdestinations promoted by New Urbanist designers(see Aurbach 2005:10-11; Dover Kohl & Partners andChael Cooper & Associates P.A. 2005:6). Clearly,people in both Brisbane and Perth are willing to walksignificant distances, especially to access rail (andbusway) services. The results also suggest that thewalk trip distances observed for certain trips inBrisbane may be larger than those observed in UScities. The mean walk distance from homes toworkplaces as part of single-mode trips in Brisbane(1.17 km) is almost triple that observed by Plaut (2005)for the US (0.32 km). And the distances walked to andfrom public transport services in Brisbane were greaterthan those observed by Hsiao et al. (1997) in OrangeCounty.

These distributions of walking distances for transportin Brisbane are current actual travel behaviour for aweekday. Of course, they do not necessarily represent

27



preference. Some of the distances walked, or even thefact that a particular person walked at all, may resultfrom that person’s travel options being constrained bythe non-availability of other travel modes. Equally,many persons may not be undertaking walk traveldue to a paucity of appropriate destinations withinwalkable distances from home, or of public transportstops that would take them to appropriatedestinations. The distributions are also silent withrespect to the quality, convenience and perceivedsafety of the walk route, and the effectiveness of anywalk/public transport interchange, and, of course,the demographic characteristics of the walkers at a

Table 3Parameters of gamma functions fitted to the distributions of walking trip distances

Distribution Gamma parameters Anderson-Darling

α β

Single-stage trips from home to allother places (Figure 2a) 1.428 0.658 4.88**

- to shops (Figure 2b) 1.332 0.608 4.42**- to primary schools (Figure 2b) 2.468 0.372 0.59- to workplaces (Figure 2b) 2.626 0.447 0.60

From home to all public transport nodes(Figure 3a) 1.459 0.523 1.49

- to bus stops (Figure 3b) 1.390 0.436 2.83*- to train stations (Figure 3b) 2.312 0.449 0.98

From all public transport nodes todestinations (Figure 4a) 1.375 0.454 10.93**

- from bus stops (Figure 4b) 1.183 0.407 8.38**- from train stations (Figure 4b) 2.126 0.370 5.30**

From any public transport node for journey to:- to shops (Figure 4c) 1.172 0.334 2.53*- to secondary schools (Figure 4c) 1.659 0.468 2.71*- to universities/TAFEs (Figure 4c) 1.533 0.497 0.77- to workplaces (Figure 4c) 1.584 0.391 3.64*

Single-stage trips between destinationsother than homes (Figure 5a) 1.286 0.521 5.01**

- between shops (Figure 5b) 1.253 0.558 2.96*- between workplaces and shops (Figure 5b) 1.592 0.330 2.33- between workplaces and restaurants/cafes (Figure 5b) 1.099 0.468 2.76*

From motor vehicle to an end destination fortrips between homes and other places (Figure 6) 0.368 6.147 52.59**

* p<0.05* * p<0.01

disaggregate level. These are topics for furtherresearch.

REFERENCESAURBACH, L (2005). TND Design Rating Standards: version2.2, Hyattsville, Maryland, USA.

BERRIGAN, D, TROIANO, RP, MCNEEL, T, DISOGRA, C,and BALLARD-BARBASH, R (2006). Active transportationincreases adherence to activity recommendations,American Journal of Preventive Medicine 31(3):210-216.

BURKE, M, and BROWN, AL (2007). How much householdtravel is ‘active transport? Unpublished manuscript.

28



PUCHER, J and DIJKSTRA, L (2003). Promoting safewalking and cycling to improve public health: lessonsfrom The Netherlands and Germany, American Journal ofPublic Health 93(9):1509.

QUEENSLAND TRANSPORT et al. (2005). South-EastQueensland Travel Survey 2003-2004: survey procedures anddocumentation – Brisbane survey, report prepared byQueensland Transport, The Urban Transport Institute, I-View, and Data Analysis Australia, Brisbane: QueenslandTransport.

STEPHENS, M (1974). EDF statistics for goodness of fit andsome comparisons, Journal of the American StatisticalAssociation 69(347):730-737.

TAYLOR, M, and YOUNG, W (1988). Traffic Analysis: newtechnology & new solutions, Hargreen Publishing Company,North Melbourne, Vic., Australia.

TRANSPORT AND POPULATION DATA CENTRE (2006).Transfigures: train access and egress modes, Transport andPopulation Data Centre, Sydney.

TUDOR-LOCKE, C, BITTMAN, M, MEROM, D, andBAUMAN, A (2005). Patterns of walking for transport andexercise: a novel application of time use data, InternationalJournal of Behavioral Nutrition and Physical Activity 2(1):5-14.

VIVIER, J (2001). Millenium Cities Database for SustainableMobility: analyses and recommendations, UITP, Brussels.

WALLACE, C (2006). The 400 metre myth – how far do peoplewalk to busway stations? Masters thesis submitted to Schoolof Built Environment, Queensland University ofTechnology.

CHALASANI, V (2005). Enriching Household TravelSurvey Data: experiences from the Microcensus 2000.Presented at the 5th Swiss Transport Research Conference, 9-11 March 2005, Monte-Verita, Ascona, Switzerland.www.strc.ch/pdf_2005/STRC05_A2_Chalasani.pdf

COLE, R, LESLIE, E, BAUMAN, A, DONALD, M andOWEN, N (2006). Socio-demographic variations in walkingfor transport for recreation or exercise among adultAustralians, Journal of Physical Activity and Health, Vol3:164-178.

DOVER KOHL & PARTNERS and CHAEL COOPER &ASSOCIATES (2005). Design Guidelines for Pedestrian-Friendly Neighbourhood Schools, City of Raleigh, NorthCarolina.

EGGER, G, DONOVAN, R, SWINBURN, B, GILES-CORTI,B, and BULL, F (1999). Physical Activity Guidelines forAustralians – Summary and Appendices, The University ofWestern Australia and the Centre for Health Promotionand Research, Sydney.

FRANK, LD (2000). Land Use and TransportationInteraction: Implications on public health and quality oflife, Journal of Planning Education and Research 20(1):6-22.

HSIAO, S, LU, J, STERLING, J, and WEATHERFORD, M(1997). Use of Geographic Information System for analysisof transit pedestrian access, Transportation Research Record1604:50-59.

JOHNSON, N, KOTZ, S and BALAKRISHNAN, N (1994).Continuous Univariate Distributions – Volume 1 (SecondEdition), Wiley-Interscience, New York.

KER, I, and GINN, S (2003). Myths and realities in walkablecatchments: the case of walking and transit, Road andTransport Research 12(2):69-80.

KRIZEK, KJ and JOHNSON, PJ (2006). Proximity to trailsand retail: effects on urban cycling and walking, Journal ofthe American Planning Association 72(1):33-42.

LAND TRANSPORT NZ (2004). Draft Pedestrian NetworkPlanning and Facilities Design Guide, Wellington, NewZealand; Land Transport NZ.

LITMAN, T (2003). Active transportation policy issues,Presented at National Roundtable on Active Transportation,9-10 April 2003, Victoria, BC, Canada.

NATIONAL INSTITUTE OF STANDARDS ANDTECHNOLOGY (NIST) and SEMATECH (2004). e-Handbookof Statistical Methods. http://www.itl.nist.gov/div898/handbook/index.htm

PLAUT, P (2005). Non-motorized commuting in the US,Transportation Research Part D: Transport and Environment10(5):347-356.

POLZIN, S and MAGGIO, E (2007). Public Transit in America:analysis of access using the 2001 National Household TravelSurvey, Tampa, Florida: Center for Urban TransportationResearch, University of South Florida.

29



Matthew Burke

Dr Matthew Burke is a trained urban planner and early career researcher whose researchinterests include travel behaviour, transport and land use integration, transport policy,and urban geography. Having previously worked in transport or land use planning atall three tiers of government in Australia, Matthew is a Research Fellow at the UrbanResearch Program, Griffith University.

Professor Lex Brown

Professor Lex Brown was trained in civil engineering, and transport and urban planning,followed by thirty years of research and teaching in the interdisciplinary environmentalfield. His work includes the transport/environment/urban planning interface,environmental assessment of projects and programs linking in to decision-making andpolicy-making arenas, environmental management in international developmentassistance, and all aspects of environmental noise.

CONTACTMatthew BurkeUrban Research ProgramGriffith University170 Kessels Road,Nathan, Qld, 4111Email: [email protected]

ACKNOWLEDGEMENTSThis research was supported under Australian Research Council’s Discovery Projectsfunding scheme (project number DP0451532). Thanks are expressed to the two refereeswho provided comments that improved this work. The authors also gratefully acknowledgeQueensland Transport for the release of the South East Queensland Travel Survey 2003/04 dataset. [The State of Queensland (Department of Transport) 2005. Note that theDepartment of Transport gives no warranty in relation to the data (including accuracy,reliability, completeness or suitability) and accepts no liability (including withoutlimitation, liability in negligence) for any loss, damage or costs (including consequentialdamage) relating to any use of the data.]

29

Distances people Abstract walk for transport South East ...

Documents

Transcript of Distances people Abstract walk for transport South East ...